CityEnergyFlow Navigator: Optimizing Building-EV Coordinate Operation on Rooftop PV Self-Consumption Performance at City Scale

Demand-side energy management centered around buildings is complex, involving rooftop photovoltaics(RPV) for power generation, energy storage devices such as batteries and ice storage systems, and the inherent flexibility of buildings themselves. In addition to traditional storage technologies, electric vehicles (EVs) are gaining attention as a new form of energy storage. However, EVs inherently exhibit uncertainties in spatiotemporal mobility and temporal variability in charging, adding greater complexity to energy management on the demand side. To address these challenges, this study proposes a city-level demand-side energy management system called CityEnergyFlow Navigator comprising the energy simulator Voyager, the visualization platform Panorama, and the experimental platform Pilot. The research focuses on the city of Shenzhen as a case study, encompassing 570,000 buildings and 470,000 EVs. Through simulations, it was found that buildings consumed 43,864,636,163 kWh, EVs consumed 712,582,586 kWh, RPV supplied 25,324,816,714 kWh, and the grid contributed 27,823,174,802 kWh of energy annually. Furthermore, the study proposes an optimization model to enhance the SCR of RPV systems through regulating the flexible resources. During the experimental period from the 116 to the 123 day, Base strategy with 470,000 EVs supporting unidirectional charging resulted in a RPV-SCR of 63.2%. However, when bidirectional charging was enabled for all EVs, the RPV-SCR increased to 76.6%. Considering the building area to ice storage energy capacity ratio of 7:1, Shenzhen city was projected to have 148,000,000 kWh ice storage units distributed among the buildings. With 470,000 EVs using default charging strategy, the RPV-SCR reached 92.3% using only these ice storage devices. Similarly, the RPV-SCR increased to 84.3% using only 28,000,000 kWh batteries. Remarkably, integrating all available flexible resources achieved an impressive RPV-SCR of 94.6%.